FeatureImpCluster: Feature Importance for Partitional Clustering

Implements a novel approach for measuring feature importance in k-means clustering. Importance of a feature is measured by the misclassification rate relative to the baseline cluster assignment due to a random permutation of feature values. An explanation of permutation feature importance in general can be found here: <https://christophm.github.io/interpretable-ml-book/feature-importance.html>.

Version: 0.1.5
Depends: data.table
Imports: ggplot2
Suggests: flexclust, clustMixType, knitr, rmarkdown, testthat, attempt, ClustImpute, covr
Published: 2021-10-20
DOI: 10.32614/CRAN.package.FeatureImpCluster
Author: Oliver Pfaffel [aut, cre]
Maintainer: Oliver Pfaffel <opfaffel at gmail.com>
License: GPL-3
NeedsCompilation: no
Materials: README NEWS
CRAN checks: FeatureImpCluster results

Documentation:

Reference manual: FeatureImpCluster.pdf

Downloads:

Package source: FeatureImpCluster_0.1.5.tar.gz
Windows binaries: r-devel: FeatureImpCluster_0.1.5.zip, r-release: FeatureImpCluster_0.1.5.zip, r-oldrel: FeatureImpCluster_0.1.5.zip
macOS binaries: r-release (arm64): FeatureImpCluster_0.1.5.tgz, r-oldrel (arm64): FeatureImpCluster_0.1.5.tgz, r-release (x86_64): FeatureImpCluster_0.1.5.tgz, r-oldrel (x86_64): FeatureImpCluster_0.1.5.tgz
Old sources: FeatureImpCluster archive

Linking:

Please use the canonical form https://CRAN.R-project.org/package=FeatureImpCluster to link to this page.